]> git.ipfire.org Git - thirdparty/systemd.git/log
thirdparty/systemd.git
4 years agopo: systemd.pot updated 18172/head
Daniele Medri [Fri, 8 Jan 2021 16:59:36 +0000 (17:59 +0100)] 
po: systemd.pot updated

4 years agopo: italian minor fix and updates
Daniele Medri [Fri, 8 Jan 2021 16:58:51 +0000 (17:58 +0100)] 
po: italian minor fix and updates

4 years agoudev/cdrom: split main() into main() and run()
Yu Watanabe [Mon, 4 Jan 2021 18:07:23 +0000 (03:07 +0900)] 
udev/cdrom: split main() into main() and run()

However, we do not use DEFINE_MAIN_FUNCTION() here, as ubuntu s390x CI
does not like it...

4 years agoudev/cdrom: drop unnecessary headers
Yu Watanabe [Mon, 28 Dec 2020 06:30:16 +0000 (15:30 +0900)] 
udev/cdrom: drop unnecessary headers

4 years agoudev/cdrom: drop never hit condition
Yu Watanabe [Mon, 28 Dec 2020 07:41:34 +0000 (16:41 +0900)] 
udev/cdrom: drop never hit condition

4 years agoudev/cdrom: void()ify several function calls
Yu Watanabe [Mon, 28 Dec 2020 06:45:54 +0000 (15:45 +0900)] 
udev/cdrom: void()ify several function calls

4 years agoudev/cdrom: tighten variable scope used in loop
Yu Watanabe [Mon, 28 Dec 2020 06:43:42 +0000 (15:43 +0900)] 
udev/cdrom: tighten variable scope used in loop

4 years agoudev/cdrom: use unaligned_read_be32() or friends
Yu Watanabe [Mon, 28 Dec 2020 06:25:19 +0000 (15:25 +0900)] 
udev/cdrom: use unaligned_read_be32() or friends

4 years agoudev/cdrom: split cd_media_info() into small pieces
Yu Watanabe [Thu, 24 Dec 2020 18:29:13 +0000 (03:29 +0900)] 
udev/cdrom: split cd_media_info() into small pieces

4 years agoudev/cdrom: introduce enum for media state
Yu Watanabe [Thu, 24 Dec 2020 17:57:37 +0000 (02:57 +0900)] 
udev/cdrom: introduce enum for media state

4 years agoudev/cdrom: move media status to Context
Yu Watanabe [Thu, 24 Dec 2020 16:00:17 +0000 (01:00 +0900)] 
udev/cdrom: move media status to Context

4 years agoudev/cdrom: introduce enum for media and drive feature
Yu Watanabe [Thu, 24 Dec 2020 09:33:54 +0000 (18:33 +0900)] 
udev/cdrom: introduce enum for media and drive feature

4 years agoudev/cdrom: introduce Context
Yu Watanabe [Thu, 24 Dec 2020 15:11:07 +0000 (00:11 +0900)] 
udev/cdrom: introduce Context

4 years agoudev/cdrom: make cd_profiles() return positive value if drive has media
Yu Watanabe [Thu, 24 Dec 2020 08:24:29 +0000 (17:24 +0900)] 
udev/cdrom: make cd_profiles() return positive value if drive has media

4 years agoudev/cdrom: make media_lock() return negative errno
Yu Watanabe [Thu, 24 Dec 2020 08:06:50 +0000 (17:06 +0900)] 
udev/cdrom: make media_lock() return negative errno

4 years agoudev/cdrom: introduce scsi_cmd_run_and_log()
Yu Watanabe [Thu, 24 Dec 2020 08:06:00 +0000 (17:06 +0900)] 
udev/cdrom: introduce scsi_cmd_run_and_log()

4 years agoudev/cdrom: use random_u64() and usleep()
Yu Watanabe [Thu, 24 Dec 2020 06:22:12 +0000 (15:22 +0900)] 
udev/cdrom: use random_u64() and usleep()

4 years agoudev/cdrom: split out parse_argv() and help()
Yu Watanabe [Thu, 24 Dec 2020 05:59:30 +0000 (14:59 +0900)] 
udev/cdrom: split out parse_argv() and help()

4 years agonetwork: route - add support to configure tcp advmss
Susant Sahani [Thu, 7 Jan 2021 15:17:17 +0000 (16:17 +0100)] 
network: route - add support to configure tcp advmss

4 years agoMerge pull request #18156 from BjoernDaase/patch-3
hadess [Fri, 8 Jan 2021 10:52:17 +0000 (11:52 +0100)] 
Merge pull request #18156 from BjoernDaase/patch-3

hwdb: Force no "mouse" type on Logitech MX Keys

4 years agohwdb: Force no "mouse" type on Logitech MX Keys 18156/head
Björn Daase [Thu, 7 Jan 2021 08:14:16 +0000 (09:14 +0100)] 
hwdb: Force no "mouse" type on Logitech MX Keys

Fixes https://gitlab.freedesktop.org/upower/upower/-/issues/128

4 years agohwdb: Relax parsing script to allow 0 and 1 for all ID_* properties
Björn Daase [Fri, 8 Jan 2021 09:32:36 +0000 (10:32 +0100)] 
hwdb: Relax parsing script to allow 0 and 1 for all ID_* properties

4 years agoman: clarify that coredumps are gc'd after 3 days
AJ Jordan [Mon, 14 Dec 2020 23:35:43 +0000 (15:35 -0800)] 
man: clarify that coredumps are gc'd after 3 days

4 years agopkg-config: make prefix overridable again
Jan Tojnar [Sat, 2 Jan 2021 01:46:33 +0000 (02:46 +0100)] 
pkg-config: make prefix overridable again

While we don't support prefix being != /usr, and this is hardcoded
all over the place, variables in pkg-config file are expected
to have overridable base directory.

This is important for at least the following two use cases:

- Installing projects to non-FHS package-specific prefixes for Nix-style
  package managers. Of course, it is then their responsibility
  to ensure systemd can find the service files.
- Installing to local path for development purposes.
  This is a compromise between running a program from a build directory,
  and running it fully installed to system prefix.

You will not want to write to system prefix in either case.

For more information, see also
https://www.bassi.io/articles/2018/03/15/pkg-config-and-paths/

Fixes https://github.com/systemd/systemd/issues/18082

Partially reverts 6e65df89c348242dbd10036abc7dd5e8181cf733

4 years agoci: turn off Azure Pipelines
Evgeny Vereshchagin [Wed, 6 Jan 2021 11:50:09 +0000 (11:50 +0000)] 
ci: turn off Azure Pipelines

4 years agomkosi: Enable InstallDirectory and SourceFileTransferFinal options
Daan De Meyer [Wed, 6 Jan 2021 17:41:44 +0000 (18:41 +0100)] 
mkosi: Enable InstallDirectory and SourceFileTransferFinal options

- InstallDirectory caches the install part of the build process
which speeds up incremental builds a little and allows inspecting
the installed components in mkosi.installdir.
- SourceFileTransferFinal copies the source files to the final
image which makes the gdb experience in qemu/systemd-nspawn a bit
nicer as it can now find the source files and show the source code
in the gdb cli itself.

4 years agostat-util: don't try to open path on path_is_temporary_fs()
Lennart Poettering [Wed, 6 Jan 2021 17:05:56 +0000 (18:05 +0100)] 
stat-util: don't try to open path on path_is_temporary_fs()

I mean, the old code at least used O_PATH, but still, we shouldn't
allocate/close an fd if we don't have to.

4 years agoMerge pull request #18044 from weblate/weblate-systemd-master
Lennart Poettering [Wed, 6 Jan 2021 19:02:54 +0000 (20:02 +0100)] 
Merge pull request #18044 from weblate/weblate-systemd-master

Translations update from Weblate

4 years agonetworkd: Enable IPv6SendRA on builtin network files
Daan De Meyer [Sat, 26 Dec 2020 14:27:55 +0000 (15:27 +0100)] 
networkd: Enable IPv6SendRA on builtin network files

RAs trigger neighbor discovery which allows users to query the
LL address of the container/VM via `ip neighbor get dev`. This is
useful as it gives users an easy way to connect to the container
without needing LLMNR or mDNS to resolve the hostname of the container
to an IP address. In practice, this allows connecting with only
networkd enabled and without resolved running in the host/container.

4 years agoMerge pull request #18150 from poettering/strextend-tweak
Lennart Poettering [Wed, 6 Jan 2021 18:55:29 +0000 (19:55 +0100)] 
Merge pull request #18150 from poettering/strextend-tweak

tree-wide: beef up strextend() a bit

4 years agosystemctl: have is-enabled return success for aliases when calling into pid1 too
Luca Boccassi [Tue, 5 Jan 2021 13:44:26 +0000 (13:44 +0000)] 
systemctl: have is-enabled return success for aliases when calling into pid1 too

commit 15d7ab87c4e5917f5788f1f8dce327a1e272bea3 introduced the
change to add an 'alias' state, but it was wired to systemctl
only when running in 'client-side' mode. Return success as
expected and documented also when running in 'server-mode'.

Fixes https://github.com/systemd/systemd/issues/18134

4 years agostring-util: use GREEDY_ALLOC_ROUND_UP() in strextend() 18150/head
Lennart Poettering [Wed, 6 Jan 2021 16:23:38 +0000 (17:23 +0100)] 
string-util: use GREEDY_ALLOC_ROUND_UP() in strextend()

This uses GREEDY_ALLOC_ROUND_UP() to grow the allocation size
exponentially. This should speed allocation loops up a bit, given
that we often call strextend() repeatedly in a loop on the same
buffer.

4 years agostring-util: imply NULL termination of strextend() argument list
Lennart Poettering [Tue, 5 Jan 2021 14:03:41 +0000 (15:03 +0100)] 
string-util: imply NULL termination of strextend() argument list

The trailing NULL in the argument list is now implied (similar to
what we already have in place in strjoin()).

4 years agoupdate TODO
Lennart Poettering [Wed, 6 Jan 2021 16:14:56 +0000 (17:14 +0100)] 
update TODO

4 years agodoc: add missing comma in DISCOVERABLE_PARTITIONS.md
Lennart Poettering [Wed, 6 Jan 2021 14:49:39 +0000 (15:49 +0100)] 
doc: add missing comma in DISCOVERABLE_PARTITIONS.md

4 years agoupdate TODO
Lennart Poettering [Mon, 28 Dec 2020 14:17:43 +0000 (15:17 +0100)] 
update TODO

4 years agoMove shared mkosi settings to a single file in mkosi.default.d/
Daan De Meyer [Sun, 3 Jan 2021 15:05:06 +0000 (16:05 +0100)] 
Move shared mkosi settings to a single file in mkosi.default.d/

4 years agocryptsetup: use crypt_token_max if available
Luca Boccassi [Mon, 28 Dec 2020 23:50:16 +0000 (23:50 +0000)] 
cryptsetup: use crypt_token_max if available

New API added upstream:

https://gitlab.com/cryptsetup/cryptsetup/-/commit/8a12f6dc2c75f8fd0c4969fbdc421895eb418072

4 years agonetwork: Allow to set device's receive queues and transmit queues
Susant Sahani [Mon, 4 Jan 2021 15:40:47 +0000 (16:40 +0100)] 
network: Allow to set device's receive queues and transmit queues

4 years agoMerge pull request #18088 from bluca/test_force_nspawn
Zbigniew Jędrzejewski-Szmek [Mon, 4 Jan 2021 19:53:38 +0000 (20:53 +0100)] 
Merge pull request #18088 from bluca/test_force_nspawn

tests: add TEST_QEMU_ONLY variable to run only tests where qemu is mandatory, TEST_PREFER_NSPAWN to run under nspawn unless unsupported, fix some testsuite issues on Debian

4 years agoUpdate 60-sensor.hwdb
milovlad [Wed, 23 Dec 2020 18:34:37 +0000 (20:34 +0200)] 
Update 60-sensor.hwdb

hwdb: Add accelerometer orientation quirk for the MYRIA MY8312

4 years agoman: use 'weak' and 'strong' for explaining difference between Wants= + Requires=
Lennart Poettering [Mon, 28 Dec 2020 12:23:37 +0000 (13:23 +0100)] 
man: use 'weak' and 'strong' for explaining difference between Wants= + Requires=

A minor tweak, that hopefully makes things a bit clearer, given that we
previously used "requirement dependency" when referring to Wants=, which
might be confusing given that we have Requires=

4 years agofix typo
Darren Ng [Mon, 4 Jan 2021 01:40:52 +0000 (09:40 +0800)] 
fix typo

4 years agoman: systemd-sysusers does not create home dir
Lucas Werkmeister [Sun, 3 Jan 2021 13:44:21 +0000 (14:44 +0100)] 
man: systemd-sysusers does not create home dir

Document that systemd-sysusers doesn’t create the specified home
directory, and point to systemd-tmpfiles as the service that can create
the home directory instead. (systemd-tmpfiles-setup.service is ordered
After=systemd-sysusers.service, so by the time systemd-tmpfiles would
create the home directory, sysusers would have created the owning user
already, and it should all work out.)

4 years agomkosi: Enable HostonlyInitrd option
Daan De Meyer [Thu, 10 Dec 2020 19:18:13 +0000 (19:18 +0000)] 
mkosi: Enable HostonlyInitrd option

Split out from #17919.

4 years agoUpdate translation files 18044/head
Weblate [Sat, 2 Jan 2021 00:37:01 +0000 (01:37 +0100)] 
Update translation files

Updated by "Update PO files to match POT (msgmerge)" hook in Weblate.

Co-authored-by: Weblate <noreply@weblate.org>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/
Translation: systemd/master

4 years agoTranslated using Weblate (French)
Julien Humbert [Sat, 2 Jan 2021 00:37:01 +0000 (01:37 +0100)] 
Translated using Weblate (French)

Currently translated at 98.3% (184 of 187 strings)

Co-authored-by: Julien Humbert <julroy67@gmail.com>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/fr/
Translation: systemd/master

4 years agoTranslated using Weblate (Romanian)
Vlad [Sat, 2 Jan 2021 00:37:01 +0000 (01:37 +0100)] 
Translated using Weblate (Romanian)

Currently translated at 100.0% (187 of 187 strings)

Co-authored-by: Vlad <milovlad@outlook.com>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/ro/
Translation: systemd/master

4 years agoTranslated using Weblate (Punjabi)
A S Alam [Sat, 2 Jan 2021 00:37:00 +0000 (01:37 +0100)] 
Translated using Weblate (Punjabi)

Currently translated at 3.2% (6 of 187 strings)

Added translation using Weblate (Punjabi)

Co-authored-by: A S Alam <amanpreet.alam@gmail.com>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/pa/
Translation: systemd/master

4 years agotests: add TEST_PREFER_NSPAWN variable to run as many as possible under nspawn 18088/head
Luca Boccassi [Thu, 31 Dec 2020 17:29:58 +0000 (17:29 +0000)] 
tests: add TEST_PREFER_NSPAWN variable to run as many as possible under nspawn

By default the test suite prefers qemu, and uses nspawn only if
a test specifically says it doesn't support qemu.
Add a variable to allow flipping the default, and run as many
tests under nspawn as possible.

4 years agotests: add TEST_QEMU_ONLY variable to run only tests where qemu is mandatory
Luca Boccassi [Sat, 26 Dec 2020 20:11:55 +0000 (20:11 +0000)] 
tests: add TEST_QEMU_ONLY variable to run only tests where qemu is mandatory

Allows to split the test run in two parts. Most tests can run under
nspawn which is much faster, and they can be ran in one chunk with
TEST_NO_QEMU=1. The qemu-only tests, which are just a handful, can
be ran in another chunk with TEST_QEMU_ONLY=1.

Allows autopkgtest to be split in two parts.

4 years agoDo not run ninja in run-integration-tests.sh
Luca Boccassi [Tue, 22 Dec 2020 17:53:14 +0000 (17:53 +0000)] 
Do not run ninja in run-integration-tests.sh

It is typically ran as root, while builds are done as normal users.
It is documented anyway to build beforehand.

4 years agotest: shortcut skip if both TEST_NO_QEMU and TEST_NO_NSPAWN are set
Luca Boccassi [Tue, 22 Dec 2020 17:51:40 +0000 (17:51 +0000)] 
test: shortcut skip if both TEST_NO_QEMU and TEST_NO_NSPAWN are set

Allows to run all tests in bulk with TEST_NO_QEMU, skipping those where it is
mandatory, without wasting time building the image.

4 years agotest: check for binariers in [usr/][s]bin
Luca Boccassi [Tue, 22 Dec 2020 17:50:34 +0000 (17:50 +0000)] 
test: check for binariers in [usr/][s]bin

The image build function greps for ExecStart lines in unit files, but some
of them (eg: systemd-firstboot) do not use a full path.
It then falls back to 'type -P' but that only works if you have the binary
installed. For optional binaries like systemd-firstboot, the installation
can then fail.

Manually check if the binary already exists in /[usr/][s]bin.

4 years agotest: use pkg-config to get user unit dir when installing dbus user socket
Luca Boccassi [Tue, 22 Dec 2020 17:47:39 +0000 (17:47 +0000)] 
test: use pkg-config to get user unit dir when installing dbus user socket

Usually on Debian ROOTLIBDIR is /lib/<arch triplet>, which is not the right place.
Use pkg-config since we define it, and then fallback to /usr/lib/systemd/user which is
the canonical location.
On both Debian&friends and Fedora dbus/dbus-broker install the user socket/service
under /usr/lib/systemd/user, not /lib/systemd/systemd/user.

4 years agotest: update test_create_image
Luca Boccassi [Tue, 22 Dec 2020 17:45:50 +0000 (17:45 +0000)] 
test: update test_create_image

Run mask_supporting_services, which was missing, and remove setup_nspawn_root which
is now already called outside test_create_image

4 years agoEnable more mkosi options
Daan De Meyer [Wed, 9 Dec 2020 21:51:56 +0000 (21:51 +0000)] 
Enable more mkosi options

- HostonlyInitrd reduces the size of the initramfs which speeds up
QEMU boot times.
- Autologin and deleting the root password remove
the manual login step whenever starting the systemd-nspawn container
or QEMU vm.
- NetworkVeth passes --network-veth to the nspawn container when
using mkosi boot and emulates --network-veth with a tap device
and networkd's 80-vm-vt.network builtin network file when using
mkosi qemu. --network-veth makes it easier to test network and
resolved in mkosi's nspawn container or QEMU vm.

4 years agoCI: Update to mkosi github action to v9
Daan De Meyer [Fri, 1 Jan 2021 14:32:52 +0000 (15:32 +0100)] 
CI: Update to mkosi github action to v9

4 years agoman: fix path reference to unit file
Gaël PORTAY [Thu, 5 Mar 2020 17:38:31 +0000 (12:38 -0500)] 
man: fix path reference to unit file

The unit files are located at path /usr/lib/systemd/system.

This fixes the path reference to the unit file by adding the missing
path component system.

4 years agodocs: fix the link to boot loader specification
Gaël PORTAY [Wed, 30 Dec 2020 14:29:36 +0000 (09:29 -0500)] 
docs: fix the link to boot loader specification

The boot loader specification link points to the boot loader interface
documentation.

This fixes the link to point to BOOT_LOADER_SPECIFICATION instead of
BOOTLOADER_INTERFACE which is itself.

4 years agoMerge pull request #18109 from yuwata/network-routing-policy-rule-family-both-fixes
Luca Boccassi [Wed, 30 Dec 2020 11:27:24 +0000 (11:27 +0000)] 
Merge pull request #18109 from yuwata/network-routing-policy-rule-family-both-fixes

network: fixes related to Family=both in [RoutingPolicyRule]

4 years agoMerge pull request #18105 from yuwata/fuzz-fix-timeout
Yu Watanabe [Wed, 30 Dec 2020 03:11:23 +0000 (12:11 +0900)] 
Merge pull request #18105 from yuwata/fuzz-fix-timeout

Fuzz fix timeout

4 years agoA typo fix in bootctl.c (#18106)
Endre Szabo [Wed, 30 Dec 2020 03:10:24 +0000 (04:10 +0100)] 
A typo fix in bootctl.c (#18106)

4 years agoresolve: slightly optimize dns_answer_add() 18105/head
Yu Watanabe [Tue, 29 Dec 2020 14:50:54 +0000 (23:50 +0900)] 
resolve: slightly optimize dns_answer_add()

Previously, dns_answer_add() was O(n^2).
With this change dns_packet_extract() becomes ~15 times faster for some
extremal case.

Before:
```
$ time ./fuzz-dns-packet ~/downloads/clusterfuzz-testcase-minimized-fuzz-dns-packet-5631106733047808
/home/watanabe/downloads/clusterfuzz-testcase-minimized-fuzz-dns-packet-5631106733047808... ok

real    0m15.453s
user    0m15.430s
sys     0m0.007s
```

After:
```
$ time ./fuzz-dns-packet ~/downloads/clusterfuzz-testcase-minimized-fuzz-dns-packet-5631106733047808
/home/watanabe/downloads/clusterfuzz-testcase-minimized-fuzz-dns-packet-5631106733047808... ok

real    0m0.831s
user    0m0.824s
sys     0m0.006s
```

Hopefully fixes oss-fuzz#19227.
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=19227

4 years agoresolve: field size in dns resource record may be zero
Yu Watanabe [Tue, 29 Dec 2020 16:48:35 +0000 (01:48 +0900)] 
resolve: field size in dns resource record may be zero

4 years agosiphash: introduce siphash24_compress_safe()
Yu Watanabe [Tue, 29 Dec 2020 16:44:04 +0000 (01:44 +0900)] 
siphash: introduce siphash24_compress_safe()

4 years agotest-network: add tests for fib rules with Family=both vs networkctl reload or reconf... 18109/head
Yu Watanabe [Tue, 29 Dec 2020 18:32:07 +0000 (03:32 +0900)] 
test-network: add tests for fib rules with Family=both vs networkctl reload or reconfigure

4 years agonetwork: drop fib rules configured with Family=both
Yu Watanabe [Tue, 29 Dec 2020 18:15:44 +0000 (03:15 +0900)] 
network: drop fib rules configured with Family=both

4 years agonetwork: set RoutingPolicyRule::family based on Family= setting
Yu Watanabe [Tue, 29 Dec 2020 18:07:33 +0000 (03:07 +0900)] 
network: set RoutingPolicyRule::family based on Family= setting

4 years agonetwork: always re-configure rules even if already exist
Yu Watanabe [Tue, 29 Dec 2020 17:45:25 +0000 (02:45 +0900)] 
network: always re-configure rules even if already exist

routing_policy_rule_get() in link_set_routing_policy_rules() does not
work when [RoutingPolicyRule] section does not have From= or To=.

4 years agonetwork: drop unnecessary checks
Yu Watanabe [Tue, 29 Dec 2020 18:17:35 +0000 (03:17 +0900)] 
network: drop unnecessary checks

By the previous commit, the checks are not necessary any more.

4 years agonetwork: fix possible memory leak
Yu Watanabe [Tue, 29 Dec 2020 17:11:32 +0000 (02:11 +0900)] 
network: fix possible memory leak

When set_put() returns 0, then already stored rule will be unref()ed
from Manager.

4 years agonetwork: constify several arguments
Yu Watanabe [Tue, 29 Dec 2020 17:03:48 +0000 (02:03 +0900)] 
network: constify several arguments

4 years agoresolve: do not compare key twice
Yu Watanabe [Tue, 29 Dec 2020 14:44:53 +0000 (23:44 +0900)] 
resolve: do not compare key twice

4 years agoresolve: slightly shorten dns_resource_key_compare_func()
Yu Watanabe [Tue, 29 Dec 2020 13:29:21 +0000 (22:29 +0900)] 
resolve: slightly shorten dns_resource_key_compare_func()

4 years agoresolve: fix typo
Yu Watanabe [Tue, 29 Dec 2020 16:22:56 +0000 (01:22 +0900)] 
resolve: fix typo

4 years agofuzzers: set maximum length for several fuzzers
Yu Watanabe [Tue, 29 Dec 2020 08:26:43 +0000 (17:26 +0900)] 
fuzzers: set maximum length for several fuzzers

Hopefully fixes oss-fuzz#19081, oss-fuzz#21731, oss-fuzz#21755.
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=19081
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=21731
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=21755

4 years agoMerge pull request #17477 from yuwata/network-drop-serialization-routing-policy-rule
Yu Watanabe [Tue, 29 Dec 2020 15:18:24 +0000 (00:18 +0900)] 
Merge pull request #17477 from yuwata/network-drop-serialization-routing-policy-rule

network: do not serialize/deserialize routing policy rules

4 years agonetwork: do not serialize/deserialize routing policy rules 17477/head
Yu Watanabe [Wed, 28 Oct 2020 08:02:41 +0000 (17:02 +0900)] 
network: do not serialize/deserialize routing policy rules

We already handle foreign routing policy rules correctly by the previous
commit. So, the serialization/deserialization of rules are not necessary
anymore.

4 years agonetwork: drop unnecessary routing policy rules
Yu Watanabe [Wed, 28 Oct 2020 11:27:23 +0000 (20:27 +0900)] 
network: drop unnecessary routing policy rules

networkd already drop foreign address, routes, and nexthops on startup,
except those created by kernel. However, previously, routing policy
rules were not. The logic of serialization/deserialization of rules only
works for rules created by previous invocation of networkd, and does not
work for one created by other tools like `ip rule`.

This makes networkd drop foreign routing policy rules except created by
kernel on startup. Also, remove rules created by networkd when the
corresponding links are dropped or networkd is stopping.

4 years agonetwork: treat rule which has l3mdev flag as created by kernel
Yu Watanabe [Thu, 29 Oct 2020 02:55:12 +0000 (11:55 +0900)] 
network: treat rule which has l3mdev flag as created by kernel

4 years agonetwork: adjust protocol of rules sent from kernel when kernel does not support FRA_P...
Yu Watanabe [Thu, 29 Oct 2020 02:34:36 +0000 (11:34 +0900)] 
network: adjust protocol of rules sent from kernel when kernel does not support FRA_PROTOCOL

Otherwise, each configured rule is treated as foreign.

4 years agonetwork: set FRA_PROTOCOL to RTPROT_STATIC by default
Yu Watanabe [Wed, 28 Oct 2020 07:54:51 +0000 (16:54 +0900)] 
network: set FRA_PROTOCOL to RTPROT_STATIC by default

4 years agoMerge pull request #18048 from poettering/timesync-man-more
Luca Boccassi [Mon, 28 Dec 2020 12:28:35 +0000 (12:28 +0000)] 
Merge pull request #18048 from poettering/timesync-man-more

man: extend time-{set,sync}.target + systemd-timesyncd/wait-sync docs

4 years agoman: apply @Minoru's suggestions from code review 18048/head
Lennart Poettering [Mon, 28 Dec 2020 09:43:10 +0000 (10:43 +0100)] 
man: apply @Minoru's suggestions from code review

Co-authored-by: Alexander Batischev <eual.jp@gmail.com>
4 years agoman: extend time-{set,sync}.target + systemd-timesyncd/wait-sync docs
Lennart Poettering [Sun, 20 Dec 2020 20:03:53 +0000 (21:03 +0100)] 
man: extend time-{set,sync}.target + systemd-timesyncd/wait-sync docs

Let's link the three man pages together more tightly and explain what
the two targets are about, emphasizing local/quick/reliable/approximate
vs remote/slow/unreliable/accurate synchronization.

Follow-up for: 1431b2f701f7ba71e5d8664e709b1fad26797918 fe934b42e480473afba8a29a4a0d3d0e789543ac

4 years agoshared/dns: fix dlopen_idn return code check
Luca Boccassi [Mon, 28 Dec 2020 00:23:14 +0000 (00:23 +0000)] 
shared/dns: fix dlopen_idn return code check

Fixes https://github.com/systemd/systemd/issues/18078

4 years agosd-dhcp-client: avoid "maybe-uninitialized" warning in client_timeout_resend()
Thomas Haller [Wed, 23 Dec 2020 12:18:56 +0000 (13:18 +0100)] 
sd-dhcp-client: avoid "maybe-uninitialized" warning in client_timeout_resend()

When compiling with CFLAGS='-Werror=maybe-uninitialized -Og' we get a
warning about uninitialized "next_timeout" variable.

Avoid the warning by adding an (unreachable) "default" label.

Fixes: c24288d21ee9 ("sd-dhcp-client: correct dhcpv4 renew/rebind retransmit timeouts")
4 years agoMerge pull request #18021 from ssahani/route-allow-list
Yu Watanabe [Thu, 24 Dec 2020 13:31:37 +0000 (22:31 +0900)] 
Merge pull request #18021 from ssahani/route-allow-list

networkd: add support for prefix allow-list and route allow-list

4 years agotest-network: add tests for [IPv6AcceptRA] PrefixDenyList= or friends 18021/head
Yu Watanabe [Thu, 24 Dec 2020 04:50:48 +0000 (13:50 +0900)] 
test-network: add tests for [IPv6AcceptRA] PrefixDenyList= or friends

4 years agonetwork: introduce RouterAllowList= and RouterDenyList= in [IPv6AcceptRA]
Yu Watanabe [Thu, 24 Dec 2020 05:27:25 +0000 (14:27 +0900)] 
network: introduce RouterAllowList= and RouterDenyList= in [IPv6AcceptRA]

4 years agonetworkd: add support for prefix allow-list and route allow-list
Susant Sahani [Fri, 18 Dec 2020 19:28:38 +0000 (20:28 +0100)] 
networkd: add support for prefix allow-list and route allow-list

4 years agonetwork: rename DenyList= -> PrefixDenyList=
Yu Watanabe [Thu, 24 Dec 2020 05:00:24 +0000 (14:00 +0900)] 
network: rename DenyList= -> PrefixDenyList=

4 years agonetwork: make RouteDenyList= filter route prefix rather than gateway address
Yu Watanabe [Thu, 24 Dec 2020 04:12:40 +0000 (13:12 +0900)] 
network: make RouteDenyList= filter route prefix rather than gateway address

DenyList= filters provided prefixes, not router address.
So, RouteDenyLisy= should so for consistency.

Fixes 16c89e649d248b0fc7c1dc7ad9f3323be4eac0f4.

4 years agonetwork: fix condition for checking the provided gateway is assigned to link
Yu Watanabe [Thu, 24 Dec 2020 04:07:30 +0000 (13:07 +0900)] 
network: fix condition for checking the provided gateway is assigned to link

Fix bug introduced by 221019166f315252304b3459902ead613b905de5.

4 years agosd-ndisc: fix indentation
Yu Watanabe [Thu, 24 Dec 2020 03:48:21 +0000 (12:48 +0900)] 
sd-ndisc: fix indentation

4 years agonetwork: drop redundant TAKE_PTR()
Yu Watanabe [Thu, 24 Dec 2020 00:13:47 +0000 (09:13 +0900)] 
network: drop redundant TAKE_PTR()

Follow-up for 16c89e649d248b0fc7c1dc7ad9f3323be4eac0f4.

4 years agoImprove instructions for debugging failing service
Sebastiaan van Stijn [Wed, 23 Dec 2020 22:51:04 +0000 (23:51 +0100)] 
Improve instructions for debugging failing service

In situations where a service fails to start, systemd suggests the user to
use "journalctl -xe" to get details about the failure. While running this
command does provide some additional details, most of the information is
similar to what was already printed when the service fails.

often the actual reason for the failure can be found in the logs of the
service that fails to start.

This patch updates the wording to suggest using "-u" to view the service
logs instead.

Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
4 years agoMerge pull request #18069 from flokli/ipv6-privacy-extensions-kernel
Yu Watanabe [Wed, 23 Dec 2020 23:56:24 +0000 (08:56 +0900)] 
Merge pull request #18069 from flokli/ipv6-privacy-extensions-kernel

network: fix IPv6PrivacyExtensions=kernel

4 years agonetwork: fix IPv6PrivacyExtensions=kernel handling 18069/head
Florian Klink [Wed, 23 Dec 2020 11:57:47 +0000 (12:57 +0100)] 
network: fix IPv6PrivacyExtensions=kernel handling

When set to "kernel", systemd is not supposed to touch that sysctl.

5e0534f1c13cd50ec2b143a8b18156cd37e502f7, part of
https://github.com/systemd/systemd/pull/17240 forgot to handle that
case.

Fixes https://github.com/systemd/systemd/issues/18003